INDEX

Explanations

terms related to measurement, legal terms, and data handling

oai_token-act-pair · gpt-4o-mini Triggered by @bot

or alternatives

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

enumii

-0.81

 يتيمه

-0.73

ंदीखरीदारी

-0.66

RegressionTest

-0.63

 utafitiHapana

-0.61

writeFieldEnd

-0.59

AutoScaleMode

-0.58

 betweenstory

-0.57

:✨

-0.57

TargetException

-0.57

POSITIVE LOGITS

or

0.51

或

0.35

0.34

 или

0.33

bzw

0.32

 hoặc

0.31

หรือ

0.29

 maupun

0.29

 oder

0.29

 또는

0.28

Activations Density 0.276%

terms related to measurement, legal terms, and data handling

or alternatives

No Comments

No Known Activations